6N1P Bass DI - Revisited

[CJ]

there is an older thread that got kind of cluttered so we are starting a new one for purposes of clarity,

here is a schematic for a Bass DI box that uses the Russian 6N1P dual triode.

these DC voltages were taken from an actual 6N1P running at 170 V-dc plate supply,

using one cathode resistor instead of two gave the measured results from the actual unit,
so we are simplfying the possibly erroneous schematic shown in the other thread,

a DIY  OPT was used,

here ya go>

 

[mrclunk]

Is there an 'off the shelf' transformer that would work? Its gapped right?

thank you btw..
Edit:
So i need a big ole SE transformer that can handle alot of DC on the primary.  Not the cheap and simple bass di as I first envisioned!

 

[My3gger]

Is there an 'off the shelf' transformer that would work? Its gapped right?

thank you btw..
Edit:
So i need a big ole SE transformer that can handle alot of DC on the primary.  Not the cheap and simple bass di as I first envisioned!

You could try Carnhill 2290 which has very close "impedances" as original. Proper datasheet will tell more about max current, although i think this is not a problem here because 8,36mA of REDDI is pretty low for saturating such core. Another model was used in SRPP version from Ian Bell.

 

[Michael Tibes]

The original transformer is from Cinemag, but I don't know the model or whether it is a custom item.

Michael

 

[mrclunk]

Thanks chaps,
I'll have a look at the data for the Carnhill 2290. It seems too good to be true tho.
My understanding is it needs high primary inductance to maintain low frequency response . I was under the impression thats why SE transformers are big and expensive?

 

[ruffrecords]

Thanks chaps,
I'll have a look at the data for the Carnhill 2290. It seems too good to be true tho.
My understanding is it needs high primary inductance to maintain low frequency response . I was under the impression thats why SE transformers are big and expensive?

I would be interested to see the data you have for the 2290. Carnhill data seems to be very sparse.

Cheers

ian

 

[mrclunk]

I would be interested to see the data you have for the 2290. Carnhill data seems to be very sparse.

Sorry i don't have any addition data to whats on the spec sheet, which is as you say sparse.
I could order one and have it sent to your lab?  If you have the time or inclination to test?

 

[PRR]

> does this not risk operating the tube close to/ beyond the dissipation limit?

No.

 

[PRR]

> I started a thread on EL34 world, simply to thank you for all your awesome input, which I noticed they deleted almost immediately

Not deleted. (That would be exceptionally odd anyway.)

"PRR shout out and +1s"
http://el34world.com/Forum/index.php?topic=19904.msg207936#msg207936

 

[ruffrecords]

Sorry i don't have any addition data to whats on the spec sheet, which is as you say sparse.
I could order one and have it sent to your lab?  If you have the time or inclination to test?

I would be happy to have a go but I am not sure if it would be very helpful. I can certainly test the primary inductance for you at 100Hz, which is as low as the meter I have goes, and it would be without dc flowing. The thing is, iron cored transformer parameters change with dc current and frequency. Inductance goes down as current goes up and inductance tends to go up as frequency goes down. Ideally it would be good to know exactly how inductance varies with both but I don't have the kit to do that. To be honest I am not even sure of the best way to do it. CJ might have some ideas.

Anyway, I am happy to measure the 100Hz inductance for you which will at least give us a starting point.

Cheers

Ian

 

[ruffrecords]

an aside question, supposing the tube DI is not connected to a circuit and the LVL is reduced, does this not risk operating the tube close to/ beyond the dissipation limit?

The quiescent current, and hence the dissipation,  is determined by the dc load line which is set by the dc  resistance of the output transformer. Since this is very low, the plate pretty much sees just the HT voltage so the dc load line is near enough a horizontal line at the HT volts. The ac load line determines the gain and it is this that is set by the reflected load.

Cheers

Ian

 

[My3gger]

"Quote from: mrclunk on November 25, 2016, 03:19:47 PM
Sorry i don't have any addition data to whats on the spec sheet, which is as you say sparse.
I could order one and have it sent to your lab?  If you have the time or inclination to test?"


I would be happy to have a go but I am not sure if it would be very helpful. I can certainly test the primary inductance for you at 100Hz, which is as low as the meter I have goes, and it would be without dc flowing. The thing is, iron cored transformer parameters change with dc current and frequency. Inductance goes down as current goes up and inductance tends to go up as frequency goes down. Ideally it would be good to know exactly how inductance varies with both but I don't have the kit to do that. To be honest I am not even sure of the best way to do it. CJ might have some ideas.

Anyway, I am happy to measure the 100Hz inductance for you which will at least give us a starting point.

Cheers

Ian

Cinemag certainly gives this kind of information, i asked them for mic input and gapped output inductance. Later at certain current and frequency for both, so why not ask Carnhill to avoid sending trafos around.
I though about having ~30mA/24V compared to ~250/5mA in core of OT. Power seems close, guess no other steps are necessary at designing them, except for lower impedances of transistors along maybe some minor things? Cores of Carnhill 2290 and custom Cinemags with iron or 50Ni/50Fe are pretty close in dimensions, as are outputs meant to be driven by transistors. 2290 expects higher impedance typical for tube output like 9600r, so it might have been requested by AML for tube projects, based on demand here.
Is my thinking correct, or expected output impedance doesn't say much about transformer quality, like those encountered from tube anodes? It seems strange Carnhill would make good low ratio and bad high ratio outputs.

 

[ruffrecords]

Cinemag certainly gives this kind of information, i asked them for mic input and gapped output inductance. Later at certain current and frequency for both, so why not ask Carnhill to avoid sending trafos around.

Good idea. We have often debayed the lack of info provided by transformer manufacturers and the best way to get it is to ask them.

I though about having ~30mA/24V compared to ~250/5mA in core of OT. Power seems close, guess no other steps are necessary at designing them, except for lower source impedances of transistors along maybe some minor things? Cores of Carnhill 2290 and custom Cinemags with iron or 50Ni/50Fe are pretty close in dimensions.

Core size is a major factor in determining power handling capability. Inductance of a gapped transformer depends a lot on dc current so higher current is likely to require a larger gap (to avoid saturation) which leads to a lower inductance. Transformer design is so cmplex it is virtually an art. I don't pretend to understand a small part of it.

2290 expects higher source impedance typical for tube output like 9600r, so it might have been requested by AML for tube projects, based on demand here.
Is my thinking correct, or expected source impedance doesn't say much about transformer quality when impedances are higher like with tubes?

A transformer does what it says on the tine; it transforms voltages, currents and impedances from one winding to another. The bandwidth over which a transformer will successfully operate is determined by the inductance, stray capacitance, leakage inductance and resistance of its windings.

For a good primer on audio transformers you should read:

http://jensen-transformers.com/wp-content/uploads/2014/09/Audio-Transformers-Chapter.pdf

Cheers

Ian

 

[My3gger]

Good idea. We have often debayed the lack of info provided by transformer manufacturers and the best way to get it is to ask them.Core size is a major factor in determining power handling capability. Inductance of a gapped transformer depends a lot on dc current so higher current is likely to require a larger gap (to avoid saturation) which leads to a lower inductance. Transformer design is so cmplex it is virtually an art. I don't pretend to understand a small part of it.A transformer does what it says on the tine; it transforms voltages, currents and impedances from one winding to another. The bandwidth over which a transformer will successfully operate is determined by the inductance, stray capacitance, leakage inductance and resistance of its windings.

For a good primer on audio transformers you should read:

http://jensen-transformers.com/wp-content/uploads/2014/09/Audio-Transformers-Chapter.pdf

Cheers

Ian

Yeah, i agree making transformers is an art. I remember talking to one of known transformer designers who told me to follow recommended output impedances, like not using 600:600 instead of 10:10k for line inputs. The same was for one version of Pultec where transformer with the same ratio, but not impedances didn't work as it should, it was Max from here who corrected me. So i'm left with Edcor 600:150, it will probably be used for EMRR's WE 141-A mash-up because output impedance will match well with primary and some step down will help too.
Wasn't one of guys making your mixer using ungapped Carnhill 2291 or another similar model? Iirc he did at least test it and later changed for Sowter. If this is so it probably can give some hints if they are properly designed for tube work, like i said manufacturer or AML should know it very well.

 

[ruffrecords]

Wasn't one of guys making your mixer using ungapped Carnhill 2291 or another similar model? Iirc he did at least test it and later changed for Sowter. If this is so it probably can give some hints if they are properly designed for tube work, like i said manufacturer or AML should know it very well.

Quite possible - I get so many questions there's no way I can remember them all. It is a good incentive to create documentation!!
You can use the 2281 (600:600) because it has a primary inductance of about 12H measured at 100Hz. This translates into a 1500 ohm load at 20Hz which is a bit on the low side for my tube designs BUT inductance tends to increase as frequency goes down so at 20Hz the inductance  could be twice this value which would make it OK. I have certainly used Sowter 600:600 transformers with no problems.

The 2291 primary inductance is around 48H so the gapped version will be less but as long as it is not far away from 12H at 100Hz then by the same reasoning it should be OK. Only way to be sure is to try it.

Cheers

Ian

 

[CJ]

inductance at the low end does not vary much with frequency in a gapped core as the perm becomes very stable with the air gap,

so your inductance reading at 100 Hz should be close to inductance at 20 Hz.

and if the core is big enough then the DC will not cause much decrease in inductance,

 

[ruffrecords]

inductance at the low end does not vary much with frequency in a gapped core as the perm becomes very stable with the air gap,

so your inductance reading at 100 Hz should be close to inductance at 20 Hz.

and if the core is big enough then the DC will not cause much decrease in inductance,

See, I said it was as much art as science! CJ has definitely put in his 10,000 hours.

Cheers

Ian

Edit: @MrClunk: Seems an inductance measurement at 100Hz will be valid. Drop me an email and I will give you my address so you can post it to me.

Cheers

Ian

 

[emrr]

Polarity looks fine, unless my quick glance mixed something. 

Ian said some version of it already: I'd change that output control so it doesn't totally short the secondary.  There's absolutely no argument for this arrangement.  Period.  The end.  Put the rheostat on the other side of the 300R's and make it 500R instead of 1K.  Or shunt a 1K across a 1K pot if you can't find a 500R. 

 

[mrclunk]

Edit: @MrClunk: Seems an inductance measurement at 100Hz will be valid. Drop me an email and I will give you my address so you can post it to me.

Thanks Ian, have just emailed AML incase they already know all this stuff. If not i'll send one your way.

 

[bancho]

Is there anybody in EU who would make this project and wants to include CJ's transformer?
Maybe we could make group order and save on the postage… postage for up to 6 is the same as for 1.
Let me know. I would take 2.